Abstract: In traditional imaging system design, optical lenses are typically optimized for artificial optimization objectives such as modulation transfer function (MTF) and field of view (FOV), resulting in structurally complex multi-element configurations. To simplify the imaging system, we propose an end-to-end collaborative design method based on optical imaging principles and use an improved genetic algorithm to design a single lens imaging system with a high-quality 40° wide-FOV. The proposed approach systematically optimizes the single lens imaging blur kernel under different FOVs, achieving: 1) Effective suppression of focal spot sidelobe energy; 2) Background stray light; 3) Improved blur kernel consistency. This significantly mitigates the non-uniform image restoration caused by spatial variations in blur kernels across different FOVs. At the same time, this method can effectively correct primary aberrations such as astigmatism and field curvature caused by large field angles. Finally, the optimized single lens system is used for real-time imaging of grayscale images and is restored through Wiener inverse filtering. A large number of comparative tests have shown that compared to traditional non spherical single lens systems, the single lens system constructed by our method improves the image quality evaluation indicators: peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM) by 11% and 15% respectively at a field of view angle of 40°, and the degree of improvement in imaging quality remains stable at around (11±3)%. This methodology can be applied to the lens design process of visible light/infrared military and civilian optoelectronic imaging equipment such as unmanned aerial vehicle monitoring, public security reconnaissance, and intelligent monitoring.